Google’s August 2026 AI Roundup: Gemini 3.7 Flash, Pixel 11, and 1B Users
Google's August 2026 AI updates: Gemini 3.7 Flash at half the price, Pixel 11 with Tensor G6, 1B Gemini users, and new video and transcription models.
Google published its August 2026 AI recap on September 1, 2026, covering a dense month of launches. The highlights: Gemini 3.7 Flash arrived three weeks after 3.6 Flash at half the token cost, four Pixel 11 devices shipped with the Tensor G6 chip, the Gemini app hit 1 billion monthly users, and new models for transcription and video generation rounded out a busy period. Here is what actually matters for developers and business operators.
What happened
| Item | Key fact |
|---|---|
| Gemini 3.7 Flash launch | Released 3 weeks after 3.6 Flash; introductory price is half the 3.6 Flash cost per million tokens |
| Pixel 11 series | Four devices (Pixel 11, 11 Pro, 11 Pro XL, 11 Pro Fold) with Tensor G6 chip running Gemini Nano |
| Gemini app users | Crossed 1 billion monthly users; Google calls it the fastest-growing product in its history |
| Gemini 3.5 Transcribe | New speech-to-text model for real-time, context-aware transcription in developer workflows |
| Gemini Omni 1.1 Flash | Video model with 4K upscaling, scene extension, and first-and-last-frame interpolation |
| Student AI plan | One year of a Google AI plan free for eligible college students worldwide |
| Daily image generation | Gemini now generates more than 150 million images per day |
Gemini 3.7 Flash: faster iteration, lower cost
Google positioned 3.7 Flash as its primary model for coding and agent workflows. It arrived just three weeks after 3.6 Flash, which is a rapid cadence, and Google says it brings “substantial improvements” across software engineering, knowledge work, and web development. The introductory price is half what 3.6 Flash cost per million tokens. For teams running high-volume agentic pipelines, that cost cut matters more than raw benchmark gains.
For context on why faster, cheaper models are reshaping how developers build, see our earlier coverage of Anthropic’s Fable 5.1 model with 75% cheaper cache reads. The trend is clear: frontier labs are competing hard on price per token, not just capability.
Pixel 11 series and Tensor G6
Google announced four devices at Made by Google 2026: the standard Pixel 11, the Pro, the Pro XL, and the Pro Fold. All four carry the Tensor G6 chip, which Google describes as its fastest and most powerful to date. The chip runs Gemini Nano on-device, which means some AI features work without a network connection. Google also flagged camera upgrades and enhanced durability across the lineup.
Gemini 3.5 Transcribe
This is a purpose-built speech-to-text model aimed at developers building voice agents, live captioning tools, and post-call analytics. Google says it converts raw audio directly into “context-aware understanding” rather than just a raw transcript, and that it handles noise and domain jargon better than conventional models. If you are building anything in the AI integration space that involves voice input, this is worth testing against Whisper and the other incumbents.
Gemini Live productivity features
Google added Personal Intelligence, Daily Brief, Spark, and hands-free inbox management to Gemini Live. The pitch is moving from a conversational assistant to one that can act on your behalf, delegating tasks by voice without you touching a screen. The feature set targets busy professionals and, according to Google’s usage data, parents in particular: busy parents are 43% more likely to use Gemini for everyday tasks than the average user.
Gemini Omni 1.1 Flash: video with more control
The new video model adds scene extension, first-and-last-frame interpolation (setting a start and end frame and letting the model fill the middle), crisp 4K upscaling, and faster prototyping. It is available in Google Flow, Google AI Studio, the Gemini Enterprise Agent Platform, and the Gemini app. Small businesses are already among the heaviest users of Gemini’s image, video, and audio creation tools for marketing materials, according to Google’s own usage data.
Why it matters
The 1 billion monthly user figure is the headline, but the operational story is the cost compression. Gemini 3.7 Flash at half the 3.6 Flash token price, arriving just three weeks later, signals that Google is willing to sacrifice margin to keep developers on its platform rather than migrating to Anthropic or OpenAI. For businesses running AI at any volume, shopping token prices every month is now a real budget lever.
On the hardware side, on-device Gemini Nano via Tensor G6 means some AI features survive in offline or low-bandwidth environments. That matters for field service businesses, logistics, or any use case where connectivity is unreliable. The Gemma open-source model for offline deployment (mentioned in the source for environments from outer space to underwater) reinforces this direction.
The student AI plan (one free year for eligible college students) is a straightforward land-and-expand move. Students who build habits around Gemini are likely to carry those into the workforce and influence purchasing decisions at employers.
Our take
Google’s August was genuinely busy, but “busy” and “impactful” are different things. The Gemini 3.7 Flash pricing cut is the most concrete win here for anyone running production AI workloads. A 50% price drop on a model that handles coding and agents well is not a press release stat, it shows up in your monthly API bill.
The 1 billion monthly users figure deserves some scrutiny. Google counts anyone who uses the Gemini app, and with it bundled across Android devices and Workspace, that number reflects distribution as much as genuine daily utility. The more revealing stat is that 63% of users now talk directly to Gemini, which suggests at least some portion of that billion are choosing voice interaction rather than just having the app installed.
For businesses thinking about workflow automation with AI, Gemini 3.7 Flash and the new transcription model are the two items worth actually testing this month. The video tools are impressive, but most business use cases for AI video are still being figured out. Start where the ROI is clearer.
What to do about it
- Check your current AI API provider’s per-token cost against Gemini 3.7 Flash’s introductory pricing. Run a like-for-like test on your actual workloads, not just benchmarks.
- If you are building voice-input features, request access to Gemini 3.5 Transcribe in Google AI Studio and compare output quality on noisy or domain-specific audio against your current solution.
- If your team uses Google Workspace, enable Gemini Live’s new productivity features (Personal Intelligence, Daily Brief) for a two-week pilot and track whether it reduces context-switching time.
- If you run a college student marketing program, check the eligibility criteria for the free one-year AI plan. Students getting free access to Gemini is a context shift for how that audience discovers AI tools.
The safest bet this month: run a cost comparison on Gemini 3.7 Flash before your next API contract renewal. That is a concrete action with a measurable result, which is more than most AI announcements offer.
Frequently asked questions
How much does Gemini 3.7 Flash cost compared to 3.6 Flash?
Google launched Gemini 3.7 Flash at an introductory price of half the cost per million tokens of Gemini 3.6 Flash.
What chip does the Pixel 11 use?
The Pixel 11 series runs on the Google Tensor G6 chip, which Google describes as its fastest and most powerful chip to date. It runs the Gemini Nano model on-device.
How many users does the Gemini app have?
The Gemini app crossed 1 billion monthly users as of August 2026, which Google says makes it the fastest-growing product in the company's history.
What is Gemini 3.5 Transcribe?
Gemini 3.5 Transcribe is Google's latest speech-to-text model designed for developer workflows such as voice agents, live captioning, and post-call analytics. It converts raw audio into context-aware transcriptions in real time.
